SEC Chair Gensler has escalated his scrutiny of AI technologies within the securities industry, warning against the practice of “AI washing” during The Messenger’s “AI: Balancing Innovation & Regulation” summit in Washington, D.C., on December 5, 2023. Concerning misrepresentations about a company’s use of AI technology and their potential, Gensler highlights that these kinds of operations fall under the same securities regulations that require “full, fair, and truthful” disclosures.
SEC scrutinizes AI implementation in financial institutions
As the SEC intensifies its focus on AI, the Division of Examinations embarks on a sweeping examination of investment advisers’ utilization and governance of AI technologies. The examination encompasses various aspects, including marketing materials, third-party providers, conflicts of interest, and contingency plans for system failure. The information gathered from this examination is poised to play a pivotal role in shaping the SEC’s rulemaking agenda regarding the regulation of AI use within the securities industry.
Despite the SEC’s acknowledgment of AI’s presence for many years and its own utilization of the technology, Chair Gensler persists in his warnings about the micro and macro issues associated with AI. The SEC’s continued scrutiny, coupled with Gensler’s apprehensions about potential biases and conflicts at a micro level and the emergence of mono cultures leading to market instability at a macro level, signifies a proactive stance toward addressing the challenges posed by AI in the financial realm.
Guidelines for regulated entities
Broker-dealers, advisers, and other regulated entities are advised to carefully navigate the complexities associated with the integration of AI technologies into their operations. Gensler emphasizes several key considerations, urging entities to understand AI models and data sets, prioritize data privacy and intellectual property protection, guard against deceptive activities, and revise disclosures and marketing materials to ensure accuracy.
Comprehend AI architectures and dataset dynamics – Regulated entities are encouraged to delve into the intricacies of AI models and the underlying data sets. Recognizing the various risks posed by algorithms and data, entities must implement robust controls, craft accurate disclosures, and adhere to relevant standards of care to maintain transparency with clients.
Safeguarding data privacy and preserving intellectual assets – The safeguarding of data privacy and intellectual property is paramount. Diligence on data sets and collection processes, particularly in generative AI scenarios, is essential. Mitigation measures include understanding data provenance, adhering to data limitation principles, deleting unnecessary data, evaluating subsets, tokenizing or anonymizing data, and human verification of AI outputs.
Addressing nonpublic material and curbing market distortions – Beyond disclosure and conflict-of-interest considerations, the SEC will closely monitor AI use for deceptive activities such as insider trading and market manipulation. Entities must scrutinize data sets for material nonpublic information, implement controls to prevent trading algorithms and generative AI from impacting securities prices artificially, and adhere to restrictions on the use or disclosure of confidential information.
Review disclosures periodically and update as needed – Industry participants are urged to continually review and update their disclosures and marketing materials. Accuracy in representations about the use of AI technologies and their capabilities is crucial for maintaining trust and regulatory compliance.
SEC and AI – Navigating Responsible Integration
As the SEC continues its comprehensive scrutiny of AI in the securities industry, the onus is on regulated entities to navigate the evolving landscape of AI integration responsibly. The challenges outlined by Chair Gensler underscore the need for a nuanced approach, balancing innovation with regulatory adherence. How will industry players respond to these guidelines, and what further measures can be taken to ensure the responsible use of AI in the ever-evolving financial landscape?
